During 1998–2013, global surface air temperature increased more slowly than greenhouse gas concentrations, and therefore this period was defined as a “global warming hiatus.” The characteristics of climate extremes in China during this period are still not clear. In this study, based on 2,841 meteorological station observations in China, the trends of precipitation and temperature extremes during the global warming hiatus were investigated relative to the reference period (RF) (1982–1997) and the whole historical period (HIS) (1982–2017). Machine learning was used to capture the nonlinear regression of climate extremes, and the Theil–Sen method with the Mann–Kendall significant test was used to calculate the slopes of their linear trends. During the global warming hiatus, annual and summer (JJA) warmest days TXx and the number of summer days SU over the whole of China and most of its four subregions were found to decrease relative to the HIS and RF periods. Annual and winter (DJF) coldest nights TNn over China and its four subregions were found to decrease moderately relative to the HIS and RF periods, whereas the number of frost days FD increased consistently. Precipitation extremes showed more temporal and spatial variability than temperature extremes. Trends of annual and JJA wet extremes during the hiatus were found to decrease relative to the HIS and RF periods, whereas the dry extreme index during the hiatus was found to increase generally over China and in most subregions. Lighter winds and lower relative humidity over most areas of China might have contributed to less pronounced trends of wet extremes during the hiatus period.